It does not block us today, as we are prototyping, but we are preparing to use this at scale over the next 2 sprints, and this would be a huge quality of life improvement.
A lot of our data in our datalake contains fp16 scalar and array columns
j
jay
10/25/2024, 5:09 AM
Let me check with the team wrt how much of a lift it might be to support it
We haven’t really prioritized it because the team is actually working on a big EOY project right now. If you’d like to take a stab we’d love a contrib!
s
Sammy Sidhu
10/25/2024, 9:25 PM
Hi @Neil Wadhvana,
Looking into this and seeing how quickly we can get this off the ground.
Can have an update for you in a day or two!
n
Neil Wadhvana
10/25/2024, 9:50 PM
Thank you @Sammy Sidhu - I appreciate the follow up